AI search glossary

A conversational search experience in Google Search where the whole result is an AI-written answer that users can ask follow-up questions about, with links to sources.

Google's AI-generated summaries that appear at the top of some search results, drawing on several web pages and linking to them as sources. Being cited in them is a core goal of GEO.

Visits that arrive from links in AI assistants' answers, identified in analytics by referrers such as chatgpt.com, perplexity.ai and gemini.google.com.

Any system that responds to a question with a written answer rather than a list of links. AI assistants like ChatGPT and Claude, and AI search features like Google AI Overviews and Perplexity, are answer engines.

The work of getting a brand named, described correctly and cited when people ask AI assistants such as ChatGPT, Gemini, Perplexity and Claude for answers or recommendations. It combines entity building, question-led content and mentions on the sources assistants trust.

See: What is AEO?, AEO services

A link to a source that an AI answer shows alongside or inside its response. Being cited sends visitors to your site and signals that the system treats your page as a trusted source.

Google's metrics for real-world page experience: Largest Contentful Paint (loading), Interaction to Next Paint (responsiveness) and Cumulative Layout Shift (visual stability).

Software that visits web pages to read and index them. Search engines and AI companies run their own, such as Googlebot, Bingbot, OAI-SearchBot, GPTBot, PerplexityBot and ClaudeBot. A site's robots.txt tells them what they may access.

Experience, Expertise, Authoritativeness and Trustworthiness: the qualities Google's quality rater guidelines use to judge whether content is reliable. Named authors, real credentials and clear sourcing all contribute.

A distinct, identifiable thing, such as a company, person, product or place, that search engines and AI systems can recognise and connect facts to. Brands that are clearly defined as entities are easier for AI systems to recommend correctly.

A highlighted answer taken from a single web page and shown near the top of Google's results. It was an early form of answer engine result and rewards the same clear, direct writing.

Optimizing content and brand signals so generative search features, such as Google AI Overviews and AI Mode, Microsoft Copilot and Perplexity, use your pages as sources and credit you in the answers they write.

See: AEO vs SEO vs GEO, GEO services

Thryv Studio's reporting model: every client gets a live workspace showing each optimization, ranking change, AI citation and lead as it happens, instead of a monthly summary.

Connecting an AI model's answer to specific sources, usually retrieved from the web or a database at the moment of answering, so the answer reflects current facts and can cite where they came from.

When an AI model states something false with confidence, such as a wrong price or a feature a product doesn't have. Clear, consistent facts across the web reduce the chance of it happening about your brand.

An open protocol that lets a website notify participating search engines, including Bing, the moment a page is added or changed, instead of waiting to be recrawled.

The format Google recommends for structured data. It sits in a script tag in the page, separate from the visible content, which makes it easy to add and maintain.

A database of entities and the relationships between them. Google's Knowledge Graph powers the information panels in search results and helps its systems understand what a brand is and how it relates to other things.

The type of AI model behind assistants such as ChatGPT, Gemini, Claude and Grok. It is trained on large amounts of text to predict and generate language, and can be connected to search to answer with current information.

A proposed Markdown file at a site's root that gives language models a short, curated map of the site's most useful pages. It is a community proposal, not an official standard.

See: llms.txt explained

Running the same set of questions (prompts) through AI assistants on a schedule and recording the answers, to measure how a brand's visibility changes over time.

A technique where an AI search system breaks one question into several related sub-questions, searches for each, and combines the results. Sites that answer the surrounding questions are more likely to be used.

A method where an AI system first retrieves relevant documents, then writes its answer using them. AI search features work this way, which is why being retrievable and quotable matters.

A plain-text file at the root of a website that tells crawlers which parts of the site they may or may not visit. It can allow or block AI crawlers individually.

Improving a website so it ranks higher in search engines' lists of results, through technical health, relevant content and authority earned from other sites. It remains the foundation that AEO and GEO build on.

See: SEO services

The percentage of AI answers to a fixed set of buyer questions that mention your brand, often compared with competitors. It is the AI-search equivalent of share of voice.

See: Measuring AI visibility

Code added to a page, usually as JSON-LD using the schema.org vocabulary, that states facts about the page in a machine-readable way: who the organization is, what a product costs, who wrote an article.

See: Schema markup for AI search

A group of pages on related subtopics that link to a central pillar page. Clusters show search engines and AI systems that a site covers a subject thoroughly.

The degree to which a site is seen as a reliable, thorough source on a subject, built by covering the subject in depth and earning references from others.

A search where the user gets what they need on the results page, from a snippet, panel or AI answer, without clicking through to any website.